(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
8 Opportunities for capturing heat
and ways to use it
Part of the energy required to convert ores to metals is required to drive chemical
reactions, but most of the rest is heat used to melt or soften the metal and to allow
diffusion. Virtually all of this heat is lost to air. It sounds as if we should try
to capture this heat, either for re-use within the same process, or by cascading
it through other industrial processes which require heat at a lower temperature.
What’s the opportunity to achieve this?
Lego® Bricks
In this chapter we’re going to play with Lego blocks—but it’s a special type of
Lego we’ve invented: each block represents a process in the long chain of processes
required to convert ores into final steel and aluminium goods. Metal in some
form flows through each block, being upgraded as it passes. There are also other
inputs to each block, and other by-products are exhausted. We could include a
long list of these other inputs and exhausts—money, energy, lubricants, labour,
chemicals and so on—but the only other inputs and exhausts we’ll consider here
are heat energy. Armed with the right box of blocks, we can now build a model
of the whole connected set of processes that interest us. In the last chapter, we
asked whether we could make any individual block more efficient. In the next one
we’re going to explore whether we may in the future invent any new blocks. In this
chapter our concern is about how they’re connected. If we connected our blocks
together in a different way, could we save significant energy? The processes in use
today have been developed independently, so would it make a big difference if we
were allowed to design them all in one go? For example, we visited a steel factory
in the North of England and watched red hot metal at around 800°C being rolled,
but then left to cool in air, even though we knew it would be reheated later on. In
Lancashire we saw aluminium cans being melted and poured into ingots which
cooled in air, so they could be shipped to Dusseldorf where they are reheated for
re-rolling. In Wales we saw scrap steel recycled in an electric arc furnace and
poured into long thin ‘blooms’, then transported two miles, and re-heated prior
to rolling. In each case it looks as if heat energy could be saved if we (had an
unlimited budget and) could reconfigure our processes. So let’s play with Lego.
8 Opportunities for capturing heat
113
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Temperature histories for case study products
If every component made of steel or aluminium is different, then every product
must be made using a different set of processes, which makes it a little difficult
to generalise in our search for opportunities to move Lego blocks around. The
only way we can begin the search is to look for representative case studies that
will illustrate the key points. We’ve done that, and Figure 8.3 on the next page
introduces our nine case study parts, split between steel and aluminium, and with
a range of different geometries and process routes.
To understand the requirements for heat energy in making these parts, we’ve
talked to all the companies involved in making them—along the journey from
ore to finished part—to obtain their temperature histories, and we’ve shown these
in the Figures 8.1 and 8.2. This data is as comprehensive as we can manage—
although different manufacturers might use slightly different temperature cycles
Hot work
Rebar
Car body
Forged mining part
Chassis plate
900
Wire
450
600
Heat treatment
Hot work
Casting
Smelting
Homogenising
Temperature (ºC)
900
Alumina production
Time
1200
114
Primary processes
0
Figure 8.1—Time/temperature
histories for steel products
Figure 8.2—Time/temperature
histories for aluminium products
Casting
1350
Heat treatment
Temperature (ºC)
1800
300
0
Time
Sustainable Materials with both eyes open
Primary processes
Extrusion
Car door
Foil
Beverage can
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
(a cycle here is a single peak on the graph—heating and cooling) for the same
product. You can see that we’ve been a little cavalier with the time axis—because
our primary concern is about the peak temperature in each cycle, rather than its
duration. On each graph the lines are all identical up to the point of casting—
because they all require the same primary processes, apart from the need to add
different alloying elements to create the right composition for each part, and this
doesn’t influence the temperature. The lines diverge from casting onwards.
Forged mining part
Car body
Forging steel allows complex, high
strength parts to be produced. A billet
is softened by heating and compressed
between shaped dies to achieve the
desired geometry. A heat treatment
consisting of quenching and tempering gives a strong and tough product.
In both steel and aluminium, car doors
have exacting requirements for both
surface quality and formability. The
surface must be free of defects in
casting and quality is improved
through subsequent hot and cold
rolling stages
Rebar
Beverage can
Steel rebar is cast as square billets
which are hot rolled to the desired bar
diameter. Strength and ductility
required are imparted by quenching
and self-tempering, where the outer
surface is cooled rapidly to form a
brittle high strength microstructure,
and tempered by the still-hot core to
restore ductility.
Aluminium beverage can bodies
require a formable sheet for drawing to
the can shape, high strength to reduce
sheet thickness and material costs, and
a high surface quality for aesthetics. Hot
and cold rolling processes give uniform
formability, while cold rolling also work
hardens the material to increase
strength.
Wire
Foil
Steel wire has very high strength and
ductility along its length. Cast billets are
hot rolled to make wire rod, with the
properties achieved by controlled
cooling followed by work hardening as
the rod is drawn to make wire.
Aluminium sheet is continuously cast
and cold rolled through multiple
passes. Annealing heat treatments are
necessary to restore ductility so that
large reductions in thickness can be
achieved.
Heavy machinery chassis plate
Extruded window frame
Plate steel is cast as thick slabs and hot
rolled to achieve the desired strength.
The plates are cut, bent and/or welded
during fabrication to produce the
finished chassis part.
Complicated profiles are produced by
extruding aluminium billets through a
shaped die. An age hardening heat
treatment increases strength.
Steel parts
Aluminium parts
Figure 8.3—Case design products
8 Opportunities for capturing heat
115
Rate of diffusion
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
0.3 Tm
Tm
Temperature (ºC)
Figure 8.4—Effect of temperature
on rate of diffusion
Yield strength (MPa)
300
225
150
75
0
100
300
500
700
Temperature (ºC)
Figure 8.5—Effect of temperature
on yield strength of an aluminium
alloy (AA6061-T6)
116
We’ll explore the processes used to make these parts in a moment, but let’s first
check that the peak temperatures in the two cycles make sense. At the Atomic
Club in chapter 5, we saw that we need energy for three reasons when making
metals: to drive chemical reactions, to allow diffusion, and to soften or melt the
metal so that forming to shape is easier. In the last chapter we saw that the chemical
reactions required to release metals from their ores occur more rapidly above their
melting temperatures which for steel and aluminium alloys are around 1500°C
and 660°C respectively. Diffusion, in which atoms move within the lattice of the
solid metal, occurs at a rate related to temperature, and may occur even at room
temperature. However the rate increases dramatically as the temperature
approaches the melting point as the graph to the side shows. Softening, the
reduction in the strength of the metal with temperature, evolves in the manner
shown in the second graph to the side. In this case, a useful reduction in strength,
say to 10 % of the cold value, occurs at around 1200°C for steel and 550°C for
aluminium. Our two temperature history graphs for our case study products show
that casting and the primary production processes occur above the melting
temperature, and subsequent processes all occur at a temperature that allows
significant diffusion, with higher temperatures when deformation is required—so
the graphs tie up with our understanding of why heat is required.
Now that we understand their temperature requirements, we can build our Lego
models of the process chains for the case study products, and on the next page
we’ve shown just two of them—for a car door (in steel) and a window frame (in
aluminium). These diagrams allow us to estimate the heat energy inputs to the
blocks. By looking at the histories and checking with the companies who do the
processing, we can also show the heat energy discarded.
On adding up the energy flows into the two chains, we can see that making the
car door requires 700 MJ, while making the window frame requires 4880 MJ.
Most of this energy is ‘lost’, in the form of heat energy discarded in exhausts,
radiated through walls, and as hot metal cools in air. We know that energy is
neither created nor destroyed, so the ‘lost’ energy cannot really be lost. Bearing
this in mind, how much of this discarded energy could we capture and re-use, to
save on the inputs?
Sustainable Materials with both eyes open
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Primary
energy
700MJ
Electricity generation
Mining
Iron &
Steelmaking
Casting
Hot rolling
Pickling
Cold rolling
Annealing/
Galvanising
Blanking &
Stamping
Losses
Figure 8.6—Energy used in
steel car door production
Primary
energy
4900MJ
Electricity generation
Bauxite
Mining
Alumina
Production
Smelting
Casting
Homogenisation
Extrusion
Finishing
Losses
4300 MJ
Figure 8.7—Energy used in aluminium
window frame production
8 Opportunities for capturing heat
117
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Exergy flows for the whole of steel and
aluminium component production
It looks as if we’ve just made a typing error in the title. “Exergy”? Surely we meant
“energy”? No, we did mean it, but “exergy”, a word invented by the splendidly
named Zoran Rant of Slovenia—is a largely unfamiliar word, but a very important
one for us now. To find out why, we’ll borrow an anecdote from our colleague Dr
Rob Miller, who teaches thermodynamics in our department:
100 litres at +5˚C
20 litres at +25˚C
Figure 8.8—Both the bath and the
shower require the same energy, but
we’d rather have the hot shower!
Let’s imagine you’re in the pub, and a dodgy character in an old coat sidles up to you
and says “I’ve got a few megajoules of heat energy in my van round the back—are
you interested?” Naturally you are—we’re all concerned about the price of heating
our homes, and keen for a good deal when some spare energy drops off the back of
a lorry, know wot I mean? But your first reaction to the offer should not be “how
much?” That way lies ruin. The right first question is “what’s its temperature?”
You should be ready to pay more for the same number of megajoules, if they’re at
a higher temperature.
The heat energy in a smaller mass of material at high temperatures is more valuable
than the same energy in a larger mass of material at lower temperature. (Figure
8.8 illustrates this with a bath-time example of exergy).This is because the higher
temperature energy can be used for heating or to generate other useful forms
of energy, such as movement. In contrast the lower temperature energy cannot
usefully be transformed or exchanged.
Heat is our main concern in this chapter, but the other forms of energy of interest
when transforming metal ores into products are:
▪▪ chemical energy that may be released during combustion of fuels;
▪▪ electrical energy in electrical currents used to drive aluminium smelting as well
as the motors and pumps used in most industrial equipment;
▪▪ mechanical energy contained in moving objects such as the rolls in rolling mills.
Could we ‘capture’ heat from the exhaust gases of furnaces, or from the hot metals
we produce, and use it—either at a lower temperature, or by converting it to one
of these three forms?
118
Sustainable Materials with both eyes open
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Let’s imagine we wanted to use the few megajoules of heat on offer in the pub to
drive a train: the question of whether we can convert heat energy into mechanical
energy depends on the temperature of the heat. We cannot avoid this, and we’ve
known about it since Nicolas Léonard Sadi Carnot, after his release from the
French army on Napoleon’s final defeat in 1815 (paving the way for the eventual
accession of his nephew Napoleon III who’s getting ready to open our next chapter),
wrote his 1824 book Reflections on the motive power of fire. Carnot showed that the
maximum work you can obtain from heat depends on the ratio (T1-T2)/T1 where
T1 is the (absolute) temperature of the heat supplied, and T2 is the temperature of
the operating environment. We can’t do much about T2, so the maximum work
depends on the temperature of our supply of heat, and therefore hotter is better.
So, although it cannot be created or destroyed, not all energy is equal: electrical
energy can be used for heating or moving, chemical energy may be used to generate
heat or electrical energy, and hotter heat energy is more useful than colder heat
energy. Zoran Rant’s term “exergy” allows us to sort this out. Exergy is defined
to be the maximum useful energy we can extract from some source of energy1. In
effect, therefore, we should be using exergy in every discussion of efficiency we ever
have—we don’t want energy efficient homes, we want them to be ‘exergy efficient’
because if we can use fuel that burns at a lower temperature to heat our living
rooms, we can save the most precious high temperature fuels for where they’re
really needed. For this reason, our first Sankey diagram in chapter 2 showing the
global transformation of energy sources into useful services was drawn using units
of exergy—the maximum work that could be obtained from each energy source
feeding our system.
And we can connect our interest in exergy in this chapter to the work of Gibbs,
whom we met in the previous chapter. Gibbs explored the fundamental limits to
energy requirements for chemical transformations from one compound to another,
now known as the Gibbs free energy. More recently, Jan Szargut a Polish engineer,
has related Gibbs free energy to a list of compounds present in the environment
to determine the standard chemical exergy of compounds. Chemical and physical
(heat) exergies are happily related and consistent: the chemical exergy is a measure
of the work required to form compounds from their natural state by separating
and reforming atomic bonds; the physical exergy is a measure of how much work
some heat at a given temperature can do while cooling to ambient temperatures.
Unlike energy, exergy is not conserved. Take for example water falling over a
waterfall, potential energy is converted firstly to kinetic energy and later to
thermal energy. Energy is conserved throughout. Yet at the bottom of the falls we
8 Opportunities for capturing heat
119
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
no longer have the option of installing a water wheel to extract useful energy—
the energy has been degraded to a lower quality and we have “lost” some exergy.
Exercising exhausts excellent exergy to external extremes? Exactly!
So exergy is the right measure of our heat flows in making steel and aluminium
components, and we can now return to our Lego models of the two process chains
from earlier. Instead of showing mass and energy flows as before, the two revised
Lego models in Figures 8.9 and 8.10 show exergy flows: the chemical exergy of
the metal, the physical exergy inputs and outputs to each Lego block, and the
lost exergy from each process. High temperature upstream processes, such as
iron-making in a blast furnace or aluminium smelting, have recoverable exergy
outputs in the form of combustible gases and hot flows of metal and exhausts.
These outputs potentially could provide a useful service. For example hot exhausts
can be used to preheat air coming into the process. In contrast, low temperature
downstream processes have little or no recoverable exergy.
In order to introduce the idea of exergy flow we’ve so far concentrated on two of
our case study products. However, in Part I of the book, we assembled enough
Exergy of
inputs
700MJ
Product,
80MJ
(12kg)
Electricity
generation
Mining
Iron &
Steelmaking
Casting
Hot rolling
Pickling
Cold rolling
Annealing/
Galvanising
Blanking &
Stamping
Yield loss,
120MJ
(18kg)
Remaining
exergy, 240MJ
Losses 190MJ
Figure 8.9—Exergy flow in
steel car door production
120
Sustainable Materials with both eyes open
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Exergy of
inputs
4900MJ
Electricity
generation
Bauxite
Mining
Product
590MJ
(20kg)
Alumina
Production
Smelting
Casting
Homogenisation
Extrusion
Finishing
Yield loss
170MJ
(6kg)
Remaining
exergy, 410MJ
Losses
3700MJ
Figure 8.10—Exergy flow in aluminium
window frame production
information to make an estimate of the exergy flows of the whole global process
of making steel and aluminium components, and these are shown in Figures 8.11
and 8.12. The chemical exergy flow looks very similar to our earlier metal flow
Sankey diagrams, because once the liquid metal has been extracted from ore, its
chemical exergy hardly changes as the geometry changes. No further chemical
reactions are involved. In addition to this flow of chemical exergy, we can also see
the exergy flows associated with fuel and electricity use throughout the process.
Had we shown energy flows rather than exergy flows, while we could see energy
being discarded from processes we wouldn’t be able to ‘value’ it—because we could
do little with it if it was at low temperature. Instead, these two diagrams show
exergy flows so the exergy discarded is, or at least could be, recovered to do useful
work. We can see that around 10 % of the output exergy in steel and aluminium
is recoverable. This is an upper estimate based on the temperatures of the flows
just as they leave the processes. Some of this output exergy is already recovered,
as we’ll see later in this chapter. The remainder of the exergy input is lost: diluted
to useless low grade heat, dissipated through furnace walls and destroyed by the
chemical reactions themselves.
8 Opportunities for capturing heat
121
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Electricity
generation
losses
10 EJ
Losses
29 EJ
Electricity
5 EJ
Off-gas
Fuel
11 EJ
Ironmaking
(BF+DRI)
Recoverable
energy 4 EJ
Steelmaking
(BOF+OH)
Coke
12 EJ
Hot
rolling
Forming
Pig iron
Liquid steel
Iron ore
Steelmaking
(EAF)
End-of-life
scrap 2 EJ
Casting
(CC+ingot)
Scrap steel
Liquid steel
Castings
(slab, billet,
bloom, ingot)
Shape casting
Cast iron
scrap
Forming scrap
Fabrication scrap
Figure 8.11—Global exergy flow for steel
122
Product
fabrication
Finished
products
Cold rolling
+coating
Construction
Vehicles
Industrial
equipment
Metal
products
Product
7 EJ
1040 Mt
The key message of these two diagrams of exergy flows is that energy used in the
downstream part of the process, supplied as electricity, does work and is converted
to low temperature heat with which we can do very little. However, the energy lost
in the earlier part of the supply chain as heat at higher temperatures has significant
remaining exergy value—and we’d like to exploit it.
Sustainable Materials with both eyes open
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Losses
4.6 EJ
Electricity
generation
losses
2.1 EJ
Smelting
Recoverable
energy 0.7 EJ
Electricty
2.0 EJ
Casting
Anode
production
Liquid aluminium
Coal
0.7 EJ
Hot
rolling
Cold
rolling
Product
fabrication
Slab
Billet
Alumina
refining
Vehicles
Extrusion
Gas
1.2 EJ
Semi-finished
products
Industrial
equipment
Construction
Oil
0.5 EJ
Metal
products
Remelting + refining
Casting
End-of-life
scrap 0.2EJ
Shape
casting
Liquid aluminium
Product
1.4 EJ
45 Mt
Casting ingot
Forming scrap
Fabrication scrap
Figure 8.12—Global exergy
flow for aluminium
So our use of exergy has revealed two opportunities: if we can cut out thermal
cycles in the processing of our components, we can reduce the need for exergy
input; where we’re discarding heat at higher temperatures, we are also discarding
useful exergy. The next two sections explore whether we can reduce the number of
thermal cycles, or we can recapture those lost exergy streams?
8 Opportunities for capturing heat
123
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Cutting out thermal cycles
Ideally we would make all steel and aluminium products with just one well
controlled thermal cycle: we’d heat up ores to extract liquid metal, adjust the
composition of the liquid, cast it, deform it to shape and provide time for required
diffusion processes, so that by the time the metal returned to ambient temperature
it was perfectly ready for use. The only immutable barrier to this ideal is that
some metallurgical treatments depend on a second thermal cycle. For example,
the processes of age hardening of aluminium alloys and tempering steel require
high temperatures for diffusion but must occur after quenching (rapid cooling) to
a lower temperature. Similarly, heat treatments such as annealing must be carried
out after cold deformation to restore ductility to the metal and allow forming.
Even in this case, the thermal cycle need not be as deep as in the graphs we
showed for our case study products—if diffusion largely stops below one third of
the melting temperature, we don’t need to cool as far as ambient temperatures.
Even in cases where we need a second thermal cycle, we’re still using more thermal
cycles than absolutely necessary, and there are three good reasons for this: we
may not have all the required equipment in the right place; it may be difficult
to co-ordinate the flow of metal through all the appropriate equipment at the
right time to catch the right temperature; some processes must be operated at
ambient temperatures. We’ll investigate these through a few examples of process
innovation.
In early steelmaking practice, the Bessemer process, or subsequently Robert
Durrer’s basic oxygen process, occurred in a separate thermal cycle from the blast
furnace. This was clearly costly, so all modern steelmaking processes are coupled:
the pig iron from the blast furnace is transferred as a liquid to the basic oxygen
furnace to avoid the extra thermal cycle. However, aluminium smelting in the
Hall-Héroult process uses a lot of electricity, so production sites have traditionally
been located near to sources of cheap electricity. These locations may be far distant
from the next process, so the aluminium is cast as 100 % pure aluminium ingots at
the smelter, transported to the site where casting will occur, and then re-melted.
Around 25 % of the world’s aluminium is re-melted in this way, for no metallurgical
benefit—just because the equipment is in different locations. At a smaller scale,
as we’ve already seen, aluminium recycling always involves ‘sweetening’ with pure
ingots, and similarly pig iron is charged as a solid into electric arc furnaces. In
both cases there is no benefit in starting from solid rather than liquid metal, and
energy would be saved by avoiding remelting if the recycling equipment could be
co-located and co-ordinated with the primary liquid metal processes.
124
Sustainable Materials with both eyes open
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
The next opportunity to cut out a thermal cycle occurs between casting and hot
rolling. We’ve seen in our Sankey diagrams of metal flow that most metal is rolled,
both to control its geometry and to break up the grain structure created by casting.
In the past in both steel and aluminium production, the liquid metal was cast as
an ingot and cooled, and then re-heated prior to hot rolling. The steel industry
has begun to move away from this practice. Instead of casting in ingots, steel
is cast in longer thinner strips, using “continuous casters”. This has the double
advantage of allowing faster cooling rates for the liquid metal, and reducing the
total amount of deformation required in subsequent rolling. The output of these
continuous casters is cut into plates, and without cooling is immediately given
some re-heating ready for rolling; this is known as hot charging, and obviously
saves the energy required to cool and reheat the cast material. A recent innovation
in Italy, described in our box story, one step beyond this hot-charging, connects
Electricity generation
Hot metal
Primary
energy
Casting
Hot rolling
Losses
Hot metal
Primary energy
Electricity generation
ESP
Figure 8.13—Comparison of Arvedi
Endless Strip Production with separate
casting and hot rolling steps
Losses
8 Opportunities for capturing heat
125
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
the continuous caster directly to the hot rolling line for production of steel strip.
This process has therefore cut out the thermal cycle between casting and hot
rolling, as shown in the Lego-block model of Figure 8.13, and as a result reduced
total requirements for energy input.
In the aluminium industry, around 30 % of the world’s sheet and foil products
(15 % of all aluminium products) are made without hot rolling, by twin-roll
casting. The box story on the next page explains the process and outlines the
benefits. However, the process is most applicable to nearly pure alloys such as foil,
because the lower alloy content gives a smaller freezing range and also because
downstream processing is not as critical for pure alloys. Directly cast strip is
more susceptible to surface defects such as porosity and ‘surface bleeds’, where
the liquid metal breaks through the thin solidified surface. These defects occur
due to the difficulty of maintaining consistent solidification with rapid cooling
and unlike conventional casting of thick slabs, after twin-roll casting there is
little opportunity for removing the surface layer or rolling. In the future, twin
roll casting may extend to low alloy content, heat treatable aluminium products
(perhaps inner panels for car bodies) and microalloyed steel products, removing
the thermal cycle involved in hot rolling and saving 2-3 GJ/t.
Arvedi Endless Strip Production7
Thin slab casting technologies link the caster and rolling mill via a soaking furnace, where the
temperature of the slab is homogenised and the production of the melt shop and rolling mill
can be separated for easier scheduling. The heat retained from casting reduces the energy
input in reheating for hot rolling. However, the largest energy savings are claimed by the
Arvedi ‘Endless Strip Production’ (ESP) process operating in Cremona in Italy, where the cast
slab is fed directly into the integrated rolling mill to produce an endless strip.
This process has a fast casting speed to achieve high productivity through the single line,
liquid core reduction and direct high reduction at caster exit to improve internal quality, and
inline induction heating for precise control of temperature.
A wide range of products may be cast and rolled through ESP, with energy savings of 1.25GJ/t
compared to reheating strip from cold. Additional benefits include reduced formation of
scale on the metal surface due to the metal spending shorter time at high temperatures,
more uniform coils as the entire strip undergoes an identical temperature and deformation
history, and lower thicknesses than can be economically achieved by conventional processes.
126
Sustainable Materials with both eyes open
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Most aluminium (the remaining 85 % of global production) is not twin-roll cast,
but cast into large ingots, typically around 2 m wide, 0.5m thick and 8m long.
Because the ingots are so large, they solidify gradually from surface to centre,
so the composition of the resulting metal changes through the thickness. The
fast cooling rate creates a different microstructure and alloy concentration at the
surfaces, so every face of the cast ingot must be removed or scalped. To allow
scalping, the ingot must be cooled to ambient temperature because we don’t yet
have cutters that operate at hotter temperatures. This is expensive because the
next process, hot rolling, operates at high temperature, so we have added an extra
thermal cycle.
(and therefore low alloy content), so is mainly used to make nonheat treatable alloys.
Twin roll casting: liquid metal
to strip in one process
Twin roll casting is the most widespread method of continuously
casting thin strips in both aluminium and steel. Liquid metal is
fed between two cooled counter-rotating rolls, with solidification
occurring on contact with the roll surfaces. Two shells form and
grow towards the roll pinch, where they are fused into sheet by a
combination of heat and pressure. Typical thicknesses are 2–4mm in
steel and 4–8mm in aluminium. The process was originally proposed
by Henry Bessemer and first commercialised by Joseph Hunter in
the 1950’s for casting aluminium strips. Today, aluminium twin roll
casters are used to produce more than 30% of all aluminium sheet
and foil products. The process works best with a short freezing range
Steel strip casting has taken longer to develop due to the higher
required process temperatures, but several plants worldwide have
demonstrated the process for low carbon and microalloyed steels5.
Twin roll casting has demonstrated large energy savings relative
to conventional routes, but there are still practical difficulties
in achieving a high quality and consistent surface finish and in
improving the lifetime of key components, particularly the casting
rolls and liquid metal containment.
Electricity generation
Primary
energy
Hot metal
Hot metal
Direct chill
casting
Sawing/
scalping
Hot rolling
Twin roll
casting
Losses
Losses
8 Opportunities for capturing heat
127
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
After casting and hot deformation, the remaining thermal cycles in our case study
products are required for heat treatments. Various innovations have aimed to
reduce the need for these cycles, particularly by integrating a heat treatment into
the period of cooling occurring after hot deformation. For the steel chassis plate
and the forged mining part among our cast studies, the steel is typically quenched
(rapidly cooled to freeze it into a strong but brittle crystal form called martensite)
and then reheated for tempering (where diffusion allows some rearrangement of
the atoms in the martensite to increase its ductility and toughness). In a clever
innovation, in producing reinforcement bars, quenching and tempering take place
in line with the rolling mill: the surface of the hot rolled bar is quenched by a
water spray, to create martensite at the surface of the bar. Sufficient heat remains
in the core of the bar that its temperature averages out to allow tempering. This
quench and self-tempering process could save about 1-1.5 GJ/t and is theoretically
possible in all cases, although it may be difficult to achieve in some forgings where
thermal stresses can cause distortion and cracking.
We’ve seen that it is possible to cut out most thermal cycles as we move towards
the ideal of only having a single thermal cycle, but many practical limits remain.
There’s also a clear commercial limit: it is easiest to create a single thermal cycle
process for one particular product—one geometry of one alloy made in high
volume. But the reality of customer needs denies this ambition. When production
chains must produce a wide variety of different products, it is more difficult to coordinate them efficiently. However ‘shorter’ production chains are possible, and
with the energy savings that may be made we should do all we can to implement
them.
Recovering and exchanging heat
Having looked at cutting out thermal cycles, what about using the heat discarded
from the various processes: can we exchange heat between one process and
another? If we don’t exchange it, can we do anything else with the heat?
Heat exchangers are familiar in daily life. The radiator in our car (at the front,
where it experiences maximum air flow) exchanges heat between the hot water
circulating round the car engine and the outside air. In turn the hot engine
exchanges heat with the cooler water leaving the radiator in order to cool the
engine. The fins at the back of our refrigerator exchange heat from the inside of
the fridge with the air in the kitchen. And the radiator that warms our living room
exchanges heat from the hot boiler with the cooler air in the room.
128
Sustainable Materials with both eyes open
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
The amount of energy transferred by a heat exchanger depends on its area (the
larger the better), the materials between which heat is transferred (liquid to liquid
or gas to gas transfer is generally better, and solid to gas or gas to solid worse), and
the temperature difference across them (the smaller the better). That last feature
causes us a problem if we try to capture and reuse heat in steel and aluminium
making: we can transfer most energy when there’s a small temperature difference,
but if we want to transfer the heat energy quickly (and therefore economically)
then we need a large temperature difference. So we must find a compromise
between quick and efficient heat recovery.
How effectively can we transfer heat between a hot gas or solid and a cool solid?
Figures 8.14 and 8.15 show the ‘recoverable exergy’ from our earlier Sankey
diagrams of global exergy flow and our options for recovering heat energy to
provide a useful service.
The hot output flows are in the form of exhaust gases, cooling liquids, waste byproducts (typically granulated solids) and the metal itself which is solid. In steel
processing, the exergy in off-gases dominates, containing approximately 80 %
of the recoverable exergy of the outputs. In aluminium processes, the heat lost
through pot walls while smelting is most significant despite being at a relatively
low temperature of around 250°C. The most common way to recover heat is to use
it to preheat inputs to furnaces (either air, fuel, or the material to be charged to
the furnace and heated) or to generate electricity. These may be combined with
cascading heat recovery (where heat recovered is used at high temperature first and
then subsequently in lower temperature processes) for further savings.
The heat in exhaust gases is transferred to air or fuel by recuperators or regenerators.
Regenerators are more suitable for higher temperature and dirtier applications
as they are less susceptible to corrosion and dirt. Incoming (solid) material may
also be preheated through direct contact with exhaust gases, for example through
aluminium stack melters, or with the Consteel® process described in the box
story on the next page. Exploiting these energy savings requires investment in
a container or conveyor for preheating, and the preheat temperature must be
controlled to avoid creation of harmful volatile organic compounds from dirty
scrap. Preheating can increase furnace productivity and may reduce metal losses
by trapping dust particles from exhaust gases and reintroducing them to the melt.
We’ve looked so far for opportunities to capture heat and re-use it within the
same industry, but potentially we could exchange heat between different
industries. This book is primarily concerned with five key materials, and like
8 Opportunities for capturing heat
129
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Hot outputs
Useful service
Thermoelectrics
Gases
Electricity
Boiler
Coke oven gas
Organic fluid
Sinter gas
Blast furnace gas
Mix
Top recovery turbine
Steam
Combustion
ORC
Kalina cycle
Rankine cycle
Steam
Basic oxygen furnace gas
Electric arc furnace exhaust
Furnace exhaust
Tunnel/basket flow
Charge preheat
Recuperator/regenerator
Gas preheat
Granulated Solids
Slag
Sinter
Dry
granulation
Air
Coke
Hot water
Wet
granulation
Low flow rate
Water
Solids
Rolled metal
Spray
cooling
Cooling
in environment
Losses
Recuperator
= 1GJ/trolled steel
Cooled
Dilution with
cold air
Furnace walls
Energy lost to environment
Figure 8.14—Exergy available in outputs
from steel production and possible
paths for waste heat recovery
130
steel and aluminium, cement requires very high temperatures (around 1450°C for
clinker production), but the paper industry requires heat at around 150-200°C
to evaporate water from wet pulp. Plastics manufacture also operates at lower
temperatures (most thermoplastics melt under 200°C). In Oxelösund and Luleå
in Sweden waste industrial heat is used to warm neighbouring houses2, and we’ve
encountered a pilot project looking at the use of waste exhaust gases to grow algae
symbiotically with steel production, where the algae also sequester a small fraction
of CO23. So, if we were given a free hand (and a huge budget) could we build an
integrated materials processing facility sharing heat among several industries, as
the industries in Kalundborg, described in the last chapter, share by-products?
Integrated thermal design is common in the chemicals industries, where most
heat transfer is from liquid to liquid, the most efficient mode for heat exchange. A
Sustainable Materials with both eyes open
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Useful service
Hot outputs
Thermoelectrics
Boiler
Organic fluid
Gases
Kalina cycle
Mix
Steam
Smelting exhaust
Reheat furnace exhaust
Charge preheat
Recuperator/regenerator
Spray
cooling
Rolled products
Steam
Rankine cycle
Stack remelter
Remelting exhaust
Solids
Electricity
ORC
Water
Gas preheat
Hot water
Recuperator
Key
Thermal exergy content
Other exergy content
Losses
Unknown exergy content
Furnace walls
Established heat path
Energy lost to environment
Figure 8.15—Exergy available in outputs
from aluminium production chain and
possible paths for waste heat recovery
Experimental/speculated
= 1GJ/trolled aluminium
technique called ‘Pinch analysis’ is often used to optimise such designs. The box
story contains some details of pinch analysis, and we anticipate that analysis of
a wider set of materials processing industries could reveal new opportunities for
heat exchange.
Finally, could we develop a heat recovery technology to exploit the heat in solid
hot metal? This could be possible with radiant heat transfer to boil a fluid, by
preheating air using convective heat transfer or by conducting heat away from
the solid surface, for example in heat pipes. Unfortunately, although our chart
shows significant exergy value in the processed metals, in practice recovering it
is difficult. Effective heat transfer requires high contact pressures, which might
damage product surfaces. Allowing the metal to cool more slowly to permit heat
exchange would allow the growth of unwanted surface oxide layers and lead to
larger than required grain sizes.
In this section we’ve seen that while there is significant exergy available in
the exhaust gases, by-products, and processed materials of both the steel and
aluminium industries, it is difficult to exploit, mainly because it is in gases or
solids and we would like to transfer it to incoming solids. As a result existing
8 Opportunities for capturing heat
131
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
practice is mainly focused on heat exchange between hot and cold gases or on the
use of exhaust gas for pre-heating scrap.
Using waste heat to generate electricity
As well as heat exchange, it’s also possible to generate electricity with waste heat
or novel thermo-electric cells, and as electricity generation itself discards waste
heat, could we usefully combine it with our other processes?
In modern power stations, the turbines are driven with steam at 500°C and a
pressure of around 30 atmospheres. Blast furnace gas cannot create these
temperatures or pressures, but recent research has shown that it can drive turbines
via steam from liquids such as benzene or ammonia instead of water. A related
development has shown that blast furnace slag can be cooled with air, rather than
water, and the resulting hot air stream can also be used to heat a working fluid.
Thermoelectric conversion offers a different approach to generating electricity
directly from heat with a solid state semiconductor converting heat flow into
electrical power. To date, commercial thermoelectric devices have low efficiencies,
around 5 %, and are very expensive. However, these efficiencies may increase, and
this approach may be able to use waste heat that cannot be exploited by any other
Energy recovery from EAF exhaust gases by Consteel®8
The Consteel® electric arc furnace (EAF) directs hot exhaust gas over an incoming conveyor
of scrap in an insulated tunnel. This warms the incoming scrap to around 300-400°C through
a combination of heat transfer and combustion of remaining carbon monoxide in the
exhaust6. The preheated scrap falls from the conveyor into a molten bath of steel within the
EAF, where it is heated further until it melts. This approach reduces the electricity needed to
heat the scrap, and savings of 0.74 GJ/t have been reported.
As well as energy savings, preheating can increase furnace productivity by reducing the time
needed for melting with a given electrical current. Metal losses in the exhaust are reduced
by trapping dust particles and reintroducing them to the melt, and as the furnace maintains
a molten bath (a ‘hot heel’), noise is reduced as there are no sparks generated as normally
occurs with a bed of solid scrap.
132
Sustainable Materials with both eyes open
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
route. For example, thermoelectric generation might be used to exploit the heat
that must be conducted through pot walls in aluminium smelting to maintain a
solid, unmelted layer to prevent corrosion of the refractory lining.
Power stations for electricity generation discard heat, and this could be used in
production processes, in an approach, generally referred to as “Combined Heat
and Power” generation or CHP. Electricity generation generally produces only low
temperature heat (up to 200°C) which is not very useful in the high temperature
steel and aluminium industries. However, in aluminium production, this heat can
be used to generate steam for the initial stages of the Bayer Process for purifying
alumina, and this application saves 15 % of current primary fuel consumption4.
Integrated steel plants have their own power stations for combusting the off gases
of primary production, and steam may also be produced for use on-site.
Pinch analysis of the steel and aluminium industries
Internal
External cooling
heat External heat input
requirement exchange requirement
800
~30%
~60%
Temperature (°C)
Cold flows
600
Hot flows
400
Minimum
temperature
difference
200
0
0
4000
8000
12000
Heat flow (kW)
16000
In the chemicals industry, pinch analysis is commonly used to derive a target for site-wide
energy consumption. This target is based on the thermodynamic maximum amount of heat
that can be recovered. Hot material flows (those at high temperature with heat available for
recovery) and cold flows (requiring heating) are surveyed and combined to generate a graph
of heat availability and demand at different temperatures. For a given minimum temperature
difference that depends on the nature of the flows (solid, liquid, gas) and the cost/area of
heat exchange, a ‘pinch point’ is defined and these composite flows will have a region of
overlap that signifies the theoretical maximum amount of heat recovery that can take place.
Outside of the overlap, the heating and cooling requirements must be supplied by external
sources; heating in furnaces and cooling in air in the case of steel and aluminium.
To achieve the targeted maximum heat recovery, heat transfer across the ‘pinch point’
temperature should be avoided. We have found that further energy savings could be
achieved beyond implementing current technologies, but that a more complicated heat
exchanger network would be needed to achieve these savings.
8 Opportunities for capturing heat
133
(C) Copyright 2012 UIT Cambridge Ltd. This electronic copy is provided, free, for personal use only.
Outlook
Steel and aluminium production requires many thermal cycles and the exhaust
and hot outputs of these processes contain valuable exergy. It is possible to reduce
the number of thermal cycles in many cases, but this may require new investment,
and could be inhibited by the need to maintain process flexibility. Heat exchange,
while theoretically very attractive, is difficult to implement because of the flows in
which heat is available and required.
Having begun this chapter with Lego, and in passing met Napoleon III who at
the time his uncle was deposed, was aged four, so presumably would have been
playing with it if only it had been invented 130 years earlier, let’s now find out how
he developed as an adult...
Notes
Exergy flows for the whole of steel and aluminium component production
1. In more detail, exergy is always defined relative to some reference state – such as ambient
temperature and pressure at sea level. Exergy is then the maximum work that can be extracted
from some source of energy while bringing the source to the same state (temperature, speed,
voltage, pressure) as the surrounding environment. In metal casting, for example, the exergy
of the liquid metal might be defined as the maximum work that can be done by the heat in the
liquid metal as it cools to room temperature.
Recovering and exchanging heat
2. SSAB, a steel producer in Sweden, supply 70% of the population of Oxelösund and Luleå with
heating using exhaust gases from their processes (SSAB, n.d).
3. Tata Steel and Sheffield University recently conducted a research project at Scunthorpe
steelworks, described by Zandi et al. (2011), where power plant exhaust gases rich in CO2
were bubbled through an algal bioreactor. The algae grow and sequester CO2 through
photosynthesis.
4. The predication of a 15% saving in primary energy by co-generation is based on research
completed for the European Union by Luo and Soria (2007).
Box stories
5. The Castrip® process operated by Nucor Steel in Crawfordsville, IN., has successfully cast
and sold steel sheet by the twin-roll casting method. The current range of grades and their
properties are documented by Sosinsky et al. (2008).
6. Memoli and Ferri (2008) describe the Consteel® technology and how both heat transfer from
the exhaust and combustion of remaining carbon monoxide in a preheat tunnel contribute to
energy savings.
Box stories
7. Image credit: Siemens press picture
8. Image reference: Tenova Consteel EAF plant
134
Sustainable Materials with both eyes open